Résumé
Recently, the use of databases for clinical trials is being promoted. We used the Japan Adult Cardiovascular Surgery Database (JACVSD) data was used as a historical control in a clinical trial, and we analyzed following : the processes of using data and the efficiency of data collection, available variables for statistical analysis, and query functions for missing and invalid data. We chose available variables of JACVSD data and created rules for merging JACVSD data with interventional group data, in addition to analyzing the data collection processes for clinical trials. Subjects were selected from cases registered in the JACVSD. On statistical analysis, 63% of 76 variables were used ; variables related to the patients' symptoms had to be collected separately. Missing and invalid data were effectively excluded. We could conduct data collection efficiently by using the JACVSD as a historical control for clinical trials. Selecting subjects from the JACVSD could reduce the burden of selecting subjects from hospitals and prevent selection bias.